{"id":82784,"date":"2026-09-24T13:50:15","date_gmt":"2026-09-24T16:50:15","guid":{"rendered":"https:\/\/a3aengenharia.com\/?post_type=articles&p=82784"},"modified":"2026-09-24T13:50:53","modified_gmt":"2026-09-24T16:50:53","slug":"chernobyl-what-happened-reactor-4-accident","status":"publish","type":"articles","link":"https:\/\/a3aengenharia.com\/en-us\/content\/technical-articles\/chernobyl-what-happened-reactor-4-accident\/","title":{"rendered":"Chernobyl: What Happened in the Reactor 4 Accident"},"content":{"rendered":"\n

In the early hours of April 26, 1986<\/strong>, Reactor 4 at the Chernobyl Nuclear Power Plant was destroyed during a test conducted before a scheduled maintenance shutdown. The test was intended to verify whether residual rotational energy from the turbine could temporarily supply certain equipment until the diesel generators came online in the event of loss of external power.<\/p>\n\n\n\n

The test began when the RBMK-1000 reactor was already in an operating condition very different from the one planned. The power reduction had been interrupted at the request of the electrical system; hours later, when the reduction resumed, power fell from approximately 500 MWt to about 30 MWt. The team managed to recover it only to around 200 MWt while xenon-135 accumulated in the core, many control rods remained far withdrawn, and the operational reactivity margin, ORM, was below the established limit.<\/p>\n\n\n\n

In this configuration, the RBMK’s large core had a power distribution that was difficult to observe and control. Steam formation in the channels reduced neutron absorption by water and could increase reactivity \u2014 the so-called positive void coefficient<\/a>. At the same time, instrumentation available at low power did not adequately show the spatial distribution of neutron flux, and ORM calculation was not presented as a simple, continuous, immediately visible safety variable for the operator.<\/p>\n\n\n\n

At 01:23:04<\/strong>, the turbine valves were closed and the test began. As the turbine-generator set slowed, some of the pumps powered by the test gradually lost speed. The combination of flow changes, water close to saturation, increased steam formation, and a positive void coefficient created feedback capable of rapidly increasing power.<\/p>\n\n\n\n

At 01:23:40<\/strong>, the AZ-5 emergency shutdown system was activated. The command should have inserted the rods and stopped the reaction. However, RBMK rods had graphite displacers and, when starting from highly withdrawn positions, their insertion initially replaced columns of water with graphite in lower regions of the core. Before the absorber material reached those regions, local positive reactivity was inserted. In a reactor that was already unstable, this effect contributed decisively to the power excursion.<\/p>\n\n\n\n

Within seconds, power increased extremely, fuel elements and pressure channels were damaged, steam generation violently raised pressure, and explosions destroyed the reactor and part of the building. The core was exposed, radioactive materials were released into the environment, and fires prolonged the release. Pripyat, the planned city a few kilometers from the plant, began to be evacuated only the following day.<\/p>\n\n\n\n

Therefore, the answer to what happened at Chernobyl<\/strong> cannot be reduced to \u201chuman error\u201d or a single defect. The accident resulted from the combination of RBMK physical characteristics, inadequate control-rod and emergency-shutdown design, insufficient instrumentation, changes to the test, operational decisions, poor communication of known risks, weak independent review, and a safety regime unable to contain production pressures.<\/p>\n\n\n\n

The INSAG-7<\/strong> report, published by the International Atomic Energy Agency in 1992, corrected important aspects of the initial 1986 interpretation. It shifted a significant portion of the emphasis that had been placed on the operators toward design deficiencies, poor safety analysis, the human-machine interface, lack of adequate operational information, regulatory weakness, and insufficient feedback from previous experience. This article serves as the hub for the technical Chernobyl learning journey<\/strong>: it presents the event broadly and directs each mechanism to a specialized chapter.<\/p>\n\n\n\n\n

What Happened at Chernobyl: Understanding the Sequence of Events<\/h2>\n\n\n\n

The essential sequence of the accident was as follows:<\/p>\n\n\n\n

    \n
  1. Reactor 4 was to be shut down for maintenance, and the team would use the shutdown to repeat an electrical test related to turbine coastdown.<\/li>\n\n\n\n
  2. The power reduction was interrupted by a request from the electrical system, prolonging operation and transferring the test to another shift.<\/li>\n\n\n\n
  3. During the later reduction, power fell from approximately 500 MWt to about 30 MWt, far below the intended level.<\/li>\n\n\n\n
  4. Xenon-135 made power recovery more difficult because it absorbed neutrons in the core.<\/li>\n\n\n\n
  5. Many rods were withdrawn to raise power, reducing the operational reactivity margin known as ORM.<\/li>\n\n\n\n
  6. The reactor was stabilized at approximately 200 MWt, and the test proceeded under a sensitive condition.<\/li>\n\n\n\n
  7. Steam formation increased reactivity because of the RBMK’s positive void coefficient.<\/li>\n\n\n\n
  8. AZ-5 was activated at 01:23:40. The rod design initially added reactivity in parts of the core before full insertion of absorber material.<\/li>\n\n\n\n
  9. Power rose rapidly, fuel channels ruptured, and the reactor was destroyed.<\/li>\n<\/ol>\n\n\n\n

    The event involved an extremely rapid power excursion, channel rupture, intense steam generation, and a pressure increase. The exact nature of the second explosion remains debated in different analyses.<\/p>\n\n\n\n\n

    Where and When Did the Accident Happen?<\/h2>\n\n\n\n

    The accident occurred in the early hours of Saturday, April 26, 1986<\/strong>, at Unit 4 of the Chernobyl Nuclear Power Plant, then located in the Ukrainian Soviet Socialist Republic. The plant was near the planned city of Pripyat and approximately 130 kilometers north of Kyiv.<\/p>\n\n\n\n

    The plant had four units in operation and two more under construction. Reactor 4, completed in 1983, was an RBMK-1000 with an electrical output of approximately 1,000 MWe and a nominal thermal power of about 3,200 MWt.<\/p>\n\n\n\n

    The industrial, political, and urban context is covered in Chernobyl Before the Accident: Plant, Pripyat, Politics, and Legacy<\/a>.<\/p>\n\n\n\n

    How Did the RBMK Reactor Work?<\/h2>\n\n\n\n

    The RBMK was a high-power Soviet reactor cooled by water, moderated by graphite, and built with numerous vertical pressure channels. Each channel housed fuel and received water circulation.<\/p>\n\n\n\n

    Uranium fission produced heat; water circulated through the channels and part of it became steam; separators supplied the turbines; and the turbines drove the electrical generators. Graphite remained in the core moderating neutrons, while water removed heat and absorbed some of them.<\/p>\n\n\n\n

    This separation contributed to a decisive characteristic: under certain configurations, more steam could mean less neutron absorption and more reactivity<\/strong>. The full architecture is described in RBMK Reactor: How the Chernobyl Reactor Worked<\/a>.<\/p>\n\n\n\n\n

    What Was the Purpose of the Test Conducted at Chernobyl?<\/h2>\n\n\n\n

    Even after the chain reaction is shut down, a nuclear reactor continues to produce heat through decay of fission products. Therefore, pumps, instrumentation, and auxiliary systems must remain powered during the transition between loss of the main source and availability of emergency power.<\/p>\n\n\n\n

    At Reactor 4, the test was intended to measure how long the turbine-generator set would continue supplying electrical voltage during coastdown. Kinetic energy stored in the rotor was expected to sustain part of the loads during the interval required for diesel generators to take over the supply. Previous tests had produced insufficient results, and modifications to the voltage-regulation system justified another attempt.<\/p>\n\n\n\n

    The need to verify this transition was technically legitimate. The problem was treating the test mainly as an electrical issue without adequate integration among those responsible for the turbine, reactor operation, and nuclear safety. The reviewed documentation indicates that the program did not clearly establish all interruption conditions, the nuclear consequences of state changes, or formal management of deviations from the approved scenario.<\/p>\n\n\n\n

    Preparation began on April 25. Reactor 4 was to be shut down for maintenance, and the power reduction was supposed to bring it to the range defined for the test. However, electrical-system dispatch requested that the unit continue generating for several hours. The reduction was interrupted, the reactor remained at roughly half power, and the test passed to another shift.<\/p>\n\n\n\n

    This delay did not by itself cause the accident, but it changed important assumptions: the state of the core, xenon buildup, the team that would execute the procedure, and pressure to complete the test before shutdown. When the reduction resumed at around 23:10, actual operation was already far from the sequence originally planned.<\/p>\n\n\n\n

    INSAG-7 also corrected the interpretation of some operational actions initially described as violations. Some conditions were permitted by procedures of the time and did not directly initiate the accident. Even so, the ease with which important functions could remain unavailable reveals a system excessively dependent on administrative controls and correct human interpretation.<\/p>\n\n\n\n

    In critical-systems engineering, a test must be treated as a temporary change to installation configuration. This requires verifiable assumptions, defined responsibilities, an interlock matrix, interruption criteria, a recovery plan, interface analysis, and recording of actual conditions. Execution should proceed only if the system remains within the analyzed envelope.<\/p>\n\n\n\n

    \n

    Engineering mitigation:<\/strong> integrated tests should not depend only on the field team’s experience. A program of testing and trials<\/a>, combined with commissioning and technical acceptance<\/a>, makes it possible to verify electrical, automation, telecommunications, protection, operation, and interruption criteria before entry into service.<\/p>\n\n\n\n

    In installations that depend on uninterrupted power and operational communications, the scope should also consider power for critical infrastructure<\/a> and a telecommunications design<\/a> compatible with supervision and emergency functions.<\/p>\n<\/div>\n\n\n\n\n

    How Did the Reactor Reach an Unstable Condition?<\/h2>\n\n\n\n

    The accident began to take shape hours before the explosions. There was no single change capable of explaining everything. Instead, there was a progressive loss of control margins.<\/p>\n\n\n\n

    The Delay in Power Reduction<\/h3>\n\n\n\n

    The reduction began on April 25 but was interrupted when load dispatch requested that the unit continue supplying power. The reactor remained for hours at approximately half power before the reduction resumed.<\/p>\n\n\n\n

    The delay changed the shift, prolonged operation, and altered the state of the core. It did not by itself cause the accident, but it became part of the sequence that moved actual execution away from the planned scenario.<\/p>\n\n\n\n

    The Drop from Approximately 500 MWt to 30 MWt<\/h3>\n\n\n\n

    At around 00:28, during transfer of power control, thermal power fell abruptly to about 30 MWt. Sources describe the exact cause with different nuances, involving possible operator action or an inadequate response by the automatic system.<\/p>\n\n\n\n

    The reactor entered a range far below the intended level. The team attempted to recover power instead of stopping the test and stabilizing the unit. Read why Reactor 4 power fell from 500 MWt to 30 MWt<\/a>.<\/p>\n\n\n\n

    Xenon Poisoning<\/h3>\n\n\n\n

    When power fell, xenon-135 absorbed neutrons and made recovery of the reaction more difficult. To compensate, control rods were progressively withdrawn.<\/p>\n\n\n\n

    Power rose to approximately 200 MWt, but the core remained in a condition of complex spatial distribution, low control margin, and high sensitivity. See iodine, xenon, and core poisoning<\/a>.<\/p>\n\n\n\n

    The Operational Reactivity Margin Was Reduced<\/h3>\n\n\n\n

    To recover power after the drop, several control rods were withdrawn. This reduced ORM, a measure that represented, in simplified terms, the reserve of neutron-absorption capability available in the core. Low ORM did not merely mean \u201cfew rods inserted\u201d: it indicated a configuration in which the reactor’s spatial behavior became more difficult to control.<\/p>\n\n\n\n

    The value was not presented to operators in an immediate and intuitive manner. Its calculation depended on the system and did not function as a simple process alarm. For a deeper explanation of this mechanism, see the article on ORM and reactivity margin<\/a>.<\/p>\n\n\n\n

    The Void Coefficient Created Positive Feedback<\/h3>\n\n\n\n

    In the RBMK, water absorbed some neutrons while graphite continued to perform the moderation function. As more water became steam, absorption decreased and reactivity could increase. More power produced more steam; more steam could produce more reactivity. Under certain conditions, this cycle became positive feedback.<\/p>\n\n\n\n

    This behavior was especially dangerous at low power and with the unfavorable configuration that had developed. See the complete explanation in the article on the positive void coefficient<\/a>.<\/p>\n\n\n\n

    The Rods Had an Initial Effect Opposite to What Was Expected<\/h3>\n\n\n\n

    The control rods contained absorber material but also graphite displacers. In certain positions, the start of insertion replaced water in the lower part of the channels with graphite before the absorber section reached that region. The result could be a local increase in reactivity during the first moments.<\/p>\n\n\n\n

    The Large Core Did Not Behave as a Single Point<\/h3>\n\n\n\n

    The RBMK core was approximately seven meters high and almost twelve meters in diameter. Distant regions could respond relatively independently, requiring control of axial and radial distribution rather than only the total power shown on the panel.<\/p>\n\n\n\n

    On the night of the accident, xenon and rod configuration contributed to an irregular distribution. This means an aggregate power value did not adequately describe the state of each region. Small local changes could rapidly shift the reaction toward more sensitive parts of the core.<\/p>\n\n\n\n

    What Could the Control Room Show?<\/h3>\n\n\n\n

    At low power, internal instrumentation did not provide the same coverage used at higher regimes. Operators depended mainly on external measurements, which were insufficient to clearly represent the spatial distribution of neutron flux.<\/p>\n\n\n\n

    ORM calculation also did not appear as a continuous and immediately understandable variable. According to INSAG-7, the system could take ten to fifteen minutes to process the measurements. In an installation with thousands of points, information existed, but it had not been transformed into situational awareness and unambiguous safety limits.<\/p>\n\n\n\n

    This is a human-machine-interface problem: alarms, trends, states, permissives, and consequences need to be presented so that the team can quickly recognize when the installation has left the analyzed envelope.<\/p>\n\n\n\n

    \n

    Mitigation through supervision and operational context:<\/strong> a SCADA<\/a> or supervision and control<\/a> solution should transform data into operating states, trends, prioritized alarms, and clear action criteria.<\/p>\n\n\n\n

    For distributed assets, remote assistance and operational monitoring<\/a>, RTUs, telemetry, synchronization, and industrial networks expand visibility, provided they are designed with availability, data quality, and clearly defined responsibilities.<\/p>\n<\/div>\n\n\n\n\n

    What Happened in the Final Seconds?<\/h2>\n\n\n\n

    The available chronology combines automatic records, process signals, operating logs, testimony, and later simulations. Some details remain debated, but the main sequence is well established.<\/p>\n\n\n\n